A few posts ago, I wrote about Aardvark.im, a site using Twitter and a few of their own proprietary algorithms to smartly route inquiries to the “resident experts” on Aardvark’s social network. There was a great article in Venture Beat discussing another Twitter-enabled tool called StockTwits.

The site is effectively a Twitter aggregator, scouring the Twittersphere for tweets with stock symbols with a dollar-sign in front of them (e.g. “$AAPL”). When you log on to the StockTwits site and give it a ticker symbol, it gives you a list of the top sixty or so users that have tweeted with that symbol in the text. Ideally, this is meant to be another channel for small investors to communicate with each other in real-time about their opinion of how fresh news headlines will affect stock prices.

Cool? Yes. But what makes it really neat is that it’s another proof-of-concept of how Twitter-based social networks can work. StockTwits collects interesting tidbits about stock information in real-time. Aardvark does something similar, and routes questions in real-time to participants in your network.

Sites like Yammer (internally-facing Twitters) are gaining traction in the corporate world (Yammer boasts Xerox and Mahalo as users), but could they be made more robust with this aggregation service?

Take a look at how prediction markets work: they’re tools to aggregate information within a community (or company). Seems like this is exactly what StockTwits’ platform does, but without the traditional dollar-associated metrics used on the popular prediction marketplaces like InTrade or Inkling (which uses play-money).

I’ll take a bet that if you could create some trust and reputation metrics around the quality of information people are tweeting about, coupled with a measure of strength-of-opinion (the size of a wager, in prediction market-speak), Twitter would essentially be a prediction market in microblogger’s clothes. Right now, the number of followers a user has is the best trust proxy the network has (which is imperfect, and readily deteriorates into a popularity contest). Perhaps a strength-of-opinion measurement could harness emerging work in semantics and artificial intelligence to gauge users’ intensity of conviction based on word choice.

One could track stocks (or more broadly, opinions about anything) with the same granularity and effectiveness seen in well-used prediction marketplaces. In fact, since Twitter sees far more traffic than any prediction market; perhaps the increased liquidity and a broad user base could translate well into even more accurate predictive muscle.

Any thoughts on uses for Twitter-based knowledge aggregation? See any cool start-ups exploring the space?